King, SL, Barton, GJ and Ranganath, LR

Interpreting sources of variation in clinical gait analysis: a case study

http://researchonline.ljmu.ac.uk/id/eprint/4815/

Article

LJMU has developed LJMU Research Online for users to access the research output of the University more effectively. Copyright © and Moral Rights for the papers on this site are retained by the individual authors and/or other copyright owners. Users may download and/or print one copy of any article(s) in LJMU Research Online to facilitate their private study or for non-commercial research. You may not engage in further distribution of the material or use it for any profit-making activities or any commercial gain.

The version presented here may differ from the published version or from the version of the record. Please see the repository URL above for details on accessing the published version and note that access may require a subscription.

For more information please contact researchonline@ljmu.ac.uk

http://researchonline.ljmu.ac.uk/

Citation (please note it is advisable to refer to the publisher’s version if you intend to cite from this work)

King, SL, Barton, GJ and Ranganath, LR (2017) Interpreting sources of variation in clinical gait analysis: a case study. Gait and Posture, 52. pp. 1-4. ISSN 0966-6362

LJMU Research Online

Interpreting sources of variation in clinical gait analysis: a case study

Short Communication

Authors and Affiliations

Stephanie L. King, PhD1, Gabor J. Barton, PhD1, Lakshminarayan R. Ranganath, MD2

1 Research Institute for Sport and Exercise Sciences, Liverpool John Moores University

2 National Alkaptonuria Centre, Royal Liverpool University Hospital, Liverpool

Corresponding author

G. J. Barton

Research Institute for Sport and Exercise Sciences,

Liverpool John Moores University,

Byrom Street,

Liverpool L3 3AF, UK

Tel: +44 (0)151 9046263,

Fax: +44 (0)151 9046284,

E-mail: g.j.barton@ljmu.ac.uk

Acknowledgements, funding sources and author contributions

We wish to thank Ms Kimberly Lewin for collecting and performing initial analyses on the

data used in this case-report.

GJB received a grant from the National Alkaptonuria Centre and SLK was employed from the

same grant income. LRR is Director of the National Alkaptonuria Centre.

All authors were fully involved in the study and preparation of the manuscript and all material

within has not been, and will not be, submitted elsewhere for publication. GJB conceived the

study and contributed to drafting the article, SLK performed the data analysis and contributed

to drafting the manuscript, and LRR contributed to critically revising the manuscript.

Highlights

Gait analysis is an important tool in identifying and evaluating movement deviations

Three forms of gait deviations exist: experimental, genuine and intentional

Gait analysts should separate these gait deviations both subjectively and objectively

A didactic description and means of interpreting gait deviations are provided

Abstract

Objective: To illustrate and discuss sources of gait deviations (experimental, genuine and

intentional) during a gait analysis and how these deviations inform clinical decision making.

Methods: A case study of a 24-year old male diagnosed with Alkaptonuria undergoing a routine

gait analysis. A 3D motion capture with the Helen-Hayes marker set was used to quantify

lower-limb joint kinematics during barefoot walking along a 10 m walkway at a self-selected

pace. Additional 2D video data were recorded in the sagittal and frontal plane. The patient

reported no aches or pains in any joint and described his lifestyle as active.

Results: Temporal-spatial parameters were within normal ranges for his age and sex. Three

sources of gait deviations were identified; the posteriorly rotated pelvis was due to an

experimental error and marker misplacement, the increased rotation of the pelvis in the

horizontal plane was genuine and observed in both 3D gait curves and in 2D video analysis,

finally the inconsistency in knee flexion/extension combined with a seemingly innocuous

interest in the consequences of abnormal gait suggested an intentional gait deviation.

Conclusions: Gait analysis is an important analytical tool in the management of a variety of

conditions that negatively impact on movement. Experienced gait analysts have the ability to

recognise genuine gait adaptations that forms part of the decision-making process for that

patient. However, their role also necessitates the ability to identify and correct for experimental

errors and critically evaluate when a deviation may not be genuine.

Keywords

Gait analysis; Alkaptonuria; osteoarthritis; malingering

Introduction

Gait analysis (GA) is becoming increasingly important and its efficacy becoming more

apparent in clinical decision making [1,2]. Gait data used to describe the movements of body

segments driven by forces, consists of 3D joint angle, moment and power curves of the ankle,

knee, hip and pelvis plotted against the gait cycle between consecutive initial contacts with the

ground. Interpretation of gait data essentially entails finding the reasons why a patient’s curves

differ from those in a normal gait database. Deviations can occur for three reasons:

experimental factors, genuine causes and intentional changes. A deviation due to experimental

error most commonly arises from inaccurate and inconsistent placement of reflective markers

over anatomical landmarks [3,4] but can also include technical aspects such as occluded

markers or tracking errors. A genuine deviation is a trusted abnormality that is a consequence

of the patient’s biomechanical or neuro-musculo-skeletal constraints and are in accord with the

patient’s clinical history. Finally, intentional, or non-habitual, deviations are most difficult to

recognise but they must be identified so that they can be ignored. A malingering patient

deliberately alters movement due to a vested interest in generating false evidence of abnormal

gait which may be of benefit to them [5].

The results of the patient presented here exhibited all three types of gait deviations. This

prompted a didactic description of how the different deviations are approached and interpreted.

Methods

Patient

A young male diagnosed with alkaptonuria (AKU) was referred for GA (further details are

withheld to prevent identification of the patient with this rare disease). Alkaptonuria is an

autosomal recessive congenital metabolic disease caused by reduced activity of an enzyme

(homogentisate 1,2-dioxygenase). This results in the accumulation of homogentisic acid

throughout the body leading to debilitating health problems including discolouration of eyes,

ears, bone and degradation of cartilage (ochronosis) ultimately resulting in severe

osteoarthritis. While AKU is present from birth, with the exception of black urine, typical

symptoms start to manifest after the mid-30s [6,7,8]. Our patient reported no aches or pains in

any joint and led an active lifestyle. Additional investigations (MRI) did not identify any lower

limb, spinal or upper limb abnormalities.

Experimental procedure

Ten Qualisys Oqus cameras (Qualisys, Gothenburg, Sweden) were used to collect kinematic

data sampling at 120 Hz. Two video cameras sampling at 25 Hz were used to collect frontal

and sagittal plane recordings. A Helen-Hayes marker set [9] was used to capture kinematic

data. The patient was instructed to walk barefoot at a self-selected pace along a 10 m walkway,

with a total of 10 trials collected. Qualisys Track Manager (Qualisys, Gothenburg, Sweden)

and Visual3D (C-Motion, Rockville, MD, USA) programs were used to track and process the

kinematic data, including cubic spline interpolation and filtering of marker positions using a

2nd order low-pass Butterworth filter with a cut-off frequency of 6 Hz. Joint angles for both

limbs were time-normalised to the gait cycle. Speed-matched reference data (N=11, age range

23-60 years) were plotted for visual comparison and coefficient of variation (CV) calculated

for key peak sagittal plane values for each joint.

It must be noted that during the experimental procedure the patient expressed interest in GA;

he specifically asked about the possible consequences and impacts on his life if an abnormal

gait was identified.

Results

Walking speed (1.30 m/s), stride length (1.46 m) and cadence (105 steps/min) were well within

normal ranges for his age and sex [10]. The CV vas markedly greater for knee flexion during

loading than any other sagittal plane peak variable (Table 1). This prompted further detailed

investigation of each walking trial to discover the source of variation. Furthermore, the visual

offset towards posterior pelvic tilt and increased range of motion in the transverse plane

prompted investigation into the observational 2D video to corroborate these findings.

Experimental deviation

The pelvic tilt is markedly lower than the expected ~10° anterior tilt that is typically seen in

young healthy individuals [11] but with relatively low CV (Figure 1 and 2a, Table 1). After

inspection of sagittal plane (side view) video data it was clear that the sacrum marker was

placed too low. This resulted in an apparent offset to posterior pelvic tilt (Figure 1 and 2a) and

a consequential apparent offset to hip extension.

Genuine deviation

The patient did not report any aches, pains or injuries therefore we did not expect many gait

deviations. However, there is a large range of motion in pelvic rotation (Figure 1 and 2b). This

deviation is consistent between repeated walks in the 3D gait report (both visually and in

reported CV) and was also visible in observational video.

Intentional deviation

The most noticeable finding in this patient’s gait was the unusual and inconsistent right knee

flexion/extension angle. This is clear both visually (Figure 1) and in the substantially large CV

(Table 1). With the exception of an extension offset during stance phase, a gait profile close to

the normal reference was achieved in some trials (Figure 1 and 2c). However the majority of

trials showed no evidence of loading response knee flexion in early stance (0-30% of gait cycle)

together with resultant plantarflexion offset of the ankle and extension offset of the hip.

Discussion

It is essential for gait analysts to correctly identify the types of deviations that can occur to

avoid interpreting non-genuine data. It is imperative to understand the technical aspects of

biomechanical analyses, particularly in cases where gait reports assist in determining treatment

interventions for patients. The typical biomechanical model used in such analyses is the Helen-

Hayes model [9] with limitations (such as cross-talk between planes at the knee and hip) widely

acknowledged [12]. Erroneous hip and pelvic curves can arise from inaccurate marker

placement of the notoriously challenging sacrum marker and the two markers on the left and

right anterior superior iliac spines (ASIS). In this patient, the lack of supporting evidence from

video of a posteriorly tilted pelvis (Figure 2a) indicates errors in marker placement. There was

no evidence of spinal deformities and he had very little excess soft tissue. As such the ASIS

markers were easily palpated and identified, therefore, by process of elimination, the offset to

posterior pelvic tilt must be due to inaccurate positioning of the sacrum marker. Mentally, gait

analysts can ‘shift’ the pelvic tilt curve and the hip flexion/extension curve upwards. However,

the same effect can be achieved in the analysis software by offsetting the 3D position of the

sacrum marker in the local reference system of the pelvis, resulting in corrected pelvic tilt

(Figure 2a) and hip flexion/extension curves. This was achieved by creating a virtual landmark,

offset from the original marker co-ordinates, and subsequently used in the computation of the

pelvis segment. We acknowledge this offset was an arbitrary correction, which in itself can

introduce some quantitative error and potentially lead to missed-diagnosis if not all sources of

information are rigorously critiqued prior to correction. However, this experimental error was

corrected for in order to align the 3D gait curves with 2D observational analysis without the

need to verify this finding in a repeat assessment.

The final gait abnormality, an intentional deviation, is more complex to identify. This requires

the ability of the skilled analyst to critically question gait data through both subjective and

objective analysis. The combination of analyses avoids any bias regarding selective

removal/evaluation of individual trials, particularly in the case of perceived intentions of the

patient. The patient was able to flex his knee during early stance in a similar pattern to speed-

matched control data during some trials (Figure 2c). However the patient had no history of

knee injury or any pain during the GA to explain this variation and apparent non-habitual gait,

and the seemingly innocuous interest in possible consequences of the GA report indicated a

potential ulterior motive. The ability to recognise such a deviation is invaluable when assessing

the functional status of patients, particularly when the information may have implications for

disease management and treatment, including potential benefit claims.

The present report highlights the three sources of gait deviations and provides didactic

examples of the integration of subjective and objective interpretations of gait curves and

observational video. The role of an experienced gait analyst is to not only recognise genuine

deviations that may require further investigation or inform a treatment plan, but also to identify

and correct for technical errors as well as have the ability to recognise when a deviation may

not be genuine.

Financial support and conflict of interests

GJB received a grant from the National Alkaptonuria Centre and SLK was employed from the

same grant income. LRR is Director of the National Alkaptonuria Centre.

Author contributions

GJB conceived the study and contributed to drafting the article, SLK performed the data

analysis and contributed to drafting the manuscript, and LRR contributed to critically revising

the manuscript.

References

1. Wren TA, Otsuka NY, Bowen RE, Scaduto AA, Chan LS, Dennis SW, et al. Outcomes

of lower extremity orthopedic surgery in ambulatory children with cerebral palsy with

and without gait analysis: results of a randomized controlled trial. Gait Posture 2013;

38:236-41.

2. Ferrari A, Brunner R, Faccioli S, Reverberi S, Benedetti MG. Gait analysis contribution

to problems identification and surgical planning in CP patients: an agreement study.

Eur J Phys Rehabil Med 2015; 51:39-48.

3. Gorton GE 3rd, Hebert DA, Gannotti ME. Assessment of the kinematic variability

among 12 motion analysis laboratories. Gait Posture 2009; 29:398-402.

4. McGinley JL, Baker R, Wolfe R, Morris ME. The reliability of three-dimensional

kinematic gait measurements: a systematic review. Gait Posture 2009; 29:360-9.

5. Benbir G, Hanagasi F, Ertan S, Hanagasi HA, Ertan T. Pure gait disturbance displayed

by malingerers: case report of two patients. Clin Neurol Neurosurg 2013; 115:1177-9.

6. Introne WJ, Gahl WA. Alkaptonuria. In: NORD guide to rare disorders. Philadelphia:

Lippincott Williams & Wilkins 2003:431.

7. Ranganath LR, Cox TF. Natural history of alkaptonuria revisited: analyses based on

scoring systems. J Inherit Metab Dis 2011; 34:1141–51.

8. Barton GJ, King SL, Robinson MA, Hawken MB, Ranganath LR. Age related deviation

of gait from normality in alkaptonuria. J Inherit Metab Dis 2015; 24: 39-44.

9. Davis RB, Ounpuu S, Tyburski D, Gage JR. A gait analysis data collection and

reduction technique. Hum Mov Sci 1991; 10:575-87.

10. Whittle M. Gait analysis: An introduction. Edinburgh: Butterworth-Heinemann.

2007:243.

11. Kirtley C. Clinical gait analysis, theory and practice. Edinburgh: Elsevier 2006:61.

12. Ferrari A, Benedetti MG, Pavan E, Frigo C, Bettinelli D, Rabuffetti M, Crenna P,

Leardini A. Quantitative comparison of five current protocols in gait analysis. Gait

Posture 2008; 28:207-16.

Table 1. Peak joint angles (Average ± SD) and coefficient of variation (CV) for key variables at the

ankle, knee, hip and pelvis.

13. Joint angle (°) CV

Peak ankle dorsiflexion 13.78 ± 0.71 5.15

Peak ankle plantarflexion -12.64 ± 2.11 16.70

Knee flexion in loading 6.68 ± 8.16 122.16

Knee flexion in swing 66.54 ± 2.53 3.80

Peak hip extension -19.75 ± 1.04 5.26

Hip flexion in loading 27.15 ± 2.35 8.65

Peak anterior pelvic tilt 4.80 ± 0.63 13.17

Peak posterior pelvic tilt 2.05 ± 0.62 30.33

Peak upwards pelvic obliquity 3.94 ± 0.53 13.50

Peak downwards pelvic obliquity -2.75 ± 0.78 28.31

Peak pelvic protraction 8.73 ± 2.03 23.24

Peak pelvic retraction -9.51 ± 2.31 24.33

Figure legends

Figure 1. A typical kinematic gait report depicting angles of the ankle, knee, hip and pelvis

(from top to bottom) in the sagittal, frontal and transverse planes (left to right). Thin black lines

with shaded area represent average (± standard deviation) patient data and thick black lines

represent speed matched control data.

Figure 2. Joint angle curves illustrating all three deviations: experimental error (top), genuine

deviation (middle) and intentional change (bottom) with our interpretation of the data. Shaded

areas represent the average (± standard deviation) for the patient and thick black lines represent

speed matched control data.

Figure 1

Ankle plantar/dorsiflexiondflx

plflx

Foot progression anglein

out

Knee flexion/extensionflex

ext

Hip flexion/extensionflex

ext

Hip abduction/adductionadd

abd

Hip internal/external rotationint

ext

Pelvic tiltant

post

Pelvic obliquityup

down

Pelvic rotationint

ext

Figure 2